bug: Upgrade Langchain, AWS Powertools, Pydantic. Fix config without …

…default embedding and step function deployment. (#598)

lib/chatbot-api/functions/api-handler/common/constant.py

@@ -5,4 +5,10 @@
 SAFE_STR_REGEX = r"^[A-Za-z0-9-_. ]*$"
 SAFE_HTTP_STR_REGEX = r"^[A-Za-z0-9-_.:/]*$"
 ID_FIELD_VALIDATION = Field(min_length=1, max_length=100, pattern=SAFE_STR_REGEX)
+ID_FIELD_VALIDATION_OPTIONAL = Field(
+    min_length=1, max_length=100, pattern=SAFE_STR_REGEX, default=None
+)
 SAFE_SHORT_STR_VALIDATION = Field(min_length=1, max_length=100, pattern=SAFE_STR_REGEX)
+SAFE_SHORT_STR_VALIDATION_OPTIONAL = Field(
+    min_length=1, max_length=100, pattern=SAFE_STR_REGEX, default=None
+)

lib/chatbot-api/functions/api-handler/index.py

@@ -52,8 +52,9 @@ def handler(event: dict, context: LambdaContext) -> dict:
         )
         return app.resolve(event, context)
     except ValidationError as e:
-        logger.warning(e.errors())
-        raise e
+        errors = e.errors(include_url=False, include_context=False, include_input=False)
+        logger.warning("Validation error", errors=errors)
+        raise ValueError(f"Invalid request. Details: {errors}")
     except CommonError as e:
         logger.warning(str(e))
         raise e

lib/chatbot-api/functions/api-handler/routes/cross_encoders.py

@@ -14,7 +14,7 @@
 
 class CrossEncodersRequest(BaseModel):
     provider: str = Field(min_length=1, max_length=500, pattern=SAFE_STR_REGEX)
-    model: str = Field(min_length=1, max_length=500, pattern=SAFE_STR_REGEX)
+    model: str = Field(min_length=1, max_length=500, pattern=r"^[A-Za-z0-9-_. /]*$")
     reference: str = Field(min_length=1, max_length=MAX_STR_INPUT_LENGTH)
     passages: List[Annotated[str, Field(min_length=1, max_length=MAX_STR_INPUT_LENGTH)]]
 

lib/chatbot-api/functions/api-handler/routes/documents.py

@@ -1,6 +1,7 @@
 import os
 from common.constant import (
     ID_FIELD_VALIDATION,
+    ID_FIELD_VALIDATION_OPTIONAL,
     SAFE_HTTP_STR_REGEX,
     SAFE_STR_REGEX,
     MAX_STR_INPUT_LENGTH,
@@ -19,9 +20,13 @@
 router = Router()
 logger = Logger()
 
+CONTENT_TYPE_VALDIATION = Field(
+    min_length=1, max_length=100, pattern=r"^[A-Za-z0-9-_./]*$"
+)
+
 
 class FileUploadRequest(BaseModel):
-    workspaceId: Optional[str] = ID_FIELD_VALIDATION
+    workspaceId: Optional[str] = ID_FIELD_VALIDATION_OPTIONAL
     fileName: str = Field(min_length=1, max_length=500, pattern=SAFE_STR_REGEX)
 
 
@@ -43,7 +48,7 @@ class WebsiteDocumentRequest(BaseModel):
     address: str = Field(min_length=1, max_length=500, pattern=SAFE_HTTP_STR_REGEX)
     followLinks: bool
     limit: int = Field(gt=-1)
-    contentTypes: Optional[List[Annotated[str, SAFE_SHORT_STR_VALIDATION]]]
+    contentTypes: Optional[List[Annotated[str, CONTENT_TYPE_VALDIATION]]] = None
 
 
 class RssFeedDocumentRequest(BaseModel):
@@ -59,16 +64,14 @@ class RssFeedDocumentRequest(BaseModel):
         default=None, min_length=1, max_length=100, pattern=SAFE_STR_REGEX
     )
     followLinks: bool
-    contentTypes: Optional[List[Annotated[str, SAFE_SHORT_STR_VALIDATION]]]
+    contentTypes: Optional[List[Annotated[str, CONTENT_TYPE_VALDIATION]]] = None
 
 
 class RssFeedCrawlerUpdateRequest(BaseModel):
     documentType: str = SAFE_SHORT_STR_VALIDATION
     followLinks: bool
     limit: int = Field(lt=500)
-    contentTypes: Optional[Annotated[str, SAFE_SHORT_STR_VALIDATION]] = Field(
-        min_length=1, max_length=100, pattern=SAFE_STR_REGEX
-    )
+    contentTypes: Optional[List[Annotated[str, CONTENT_TYPE_VALDIATION]]] = None
 
 
 class ListDocumentsRequest(BaseModel):

lib/chatbot-api/functions/api-handler/routes/embeddings.py

@@ -15,7 +15,7 @@
 
 class EmbeddingsRequest(BaseModel):
     provider: str = Field(min_length=1, max_length=500, pattern=SAFE_STR_REGEX)
-    model: str = Field(min_length=1, max_length=500, pattern=SAFE_STR_REGEX)
+    model: str = Field(min_length=1, max_length=500, pattern=r"^[A-Za-z0-9-_. /]*$")
     passages: List[Annotated[str, Field(min_length=1, max_length=MAX_STR_INPUT_LENGTH)]]
     task: Optional[Task] = Task.STORE
 

lib/chatbot-api/functions/api-handler/routes/workspaces.py

@@ -1,6 +1,7 @@
 from typing import Annotated, List, Optional
 from common.constant import (
     SAFE_SHORT_STR_VALIDATION,
+    SAFE_SHORT_STR_VALIDATION_OPTIONAL,
 )
 from common.validation import WorkspaceIdValidation
 import genai_core.types
@@ -27,9 +28,13 @@ class CreateWorkspaceAuroraRequest(BaseModel):
     kind: str = SAFE_SHORT_STR_VALIDATION
     name: str = Field(min_length=1, max_length=100, pattern=name_regex)
     embeddingsModelProvider: str = SAFE_SHORT_STR_VALIDATION
-    embeddingsModelName: str = SAFE_SHORT_STR_VALIDATION
-    crossEncoderModelProvider: Optional[str] = SAFE_SHORT_STR_VALIDATION
-    crossEncoderModelName: Optional[str] = SAFE_SHORT_STR_VALIDATION
+    embeddingsModelName: str = Field(
+        min_length=0, max_length=500, pattern=r"^[A-Za-z0-9-_. /]*$", default=None
+    )
+    crossEncoderModelProvider: Optional[str] = SAFE_SHORT_STR_VALIDATION_OPTIONAL
+    crossEncoderModelName: Optional[str] = Field(
+        min_length=0, max_length=500, pattern=r"^[A-Za-z0-9-_. /]*$", default=None
+    )
     languages: List[Annotated[str, SAFE_SHORT_STR_VALIDATION]]
     metric: str = SAFE_SHORT_STR_VALIDATION
     index: bool

lib/chatbot-api/functions/resolvers/send-query-lambda-resolver/index.py

@@ -6,7 +6,8 @@
 from aws_lambda_powertools import Logger, Tracer
 from aws_lambda_powertools.utilities.typing import LambdaContext
 from aws_lambda_powertools.logging import correlation_paths
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, ValidationError
+
 
 tracer = Tracer()
 logger = Logger(log_uncaught_exceptions=True)
@@ -17,29 +18,36 @@
 MAX_STR_INPUT_LENGTH = 1000000
 SAFE_STR_REGEX = r"^[A-Za-z0-9-_. ]*$"
 SAFE_SHORT_STR_VALIDATION = Field(min_length=0, max_length=500, pattern=SAFE_STR_REGEX)
+SAFE_SHORT_STR_VALIDATION_OPTIONAL = Field(
+    min_length=0, max_length=500, pattern=SAFE_STR_REGEX, default=None
+)
 
 
 class ModelKwargsFieldValidation(BaseModel):
-    streaming: Optional[bool]
-    maxTokens: Optional[int] = Field(gt=0, lt=1000000)
-    temperature: Optional[float] = Field(ge=0, le=1)
-    topP: Optional[float] = Field(ge=0, le=1)
+    streaming: Optional[bool] = None
+    maxTokens: Optional[int] = Field(gt=0, lt=1000000, default=None)
+    temperature: Optional[float] = Field(ge=0, le=1, default=None)
+    topP: Optional[float] = Field(ge=0, le=1, default=None)
 
 
 class FileFieldValidation(BaseModel):
-    provider: Optional[str] = SAFE_SHORT_STR_VALIDATION
-    key: Optional[str] = SAFE_SHORT_STR_VALIDATION
+    provider: Optional[str] = SAFE_SHORT_STR_VALIDATION_OPTIONAL
+    key: Optional[str] = SAFE_SHORT_STR_VALIDATION_OPTIONAL
 
 
 class DataFieldValidation(BaseModel):
-    modelName: Optional[str] = SAFE_SHORT_STR_VALIDATION
-    provider: Optional[str] = SAFE_SHORT_STR_VALIDATION
-    sessionId: Optional[str] = SAFE_SHORT_STR_VALIDATION
-    workspaceId: Optional[str] = SAFE_SHORT_STR_VALIDATION
-    mode: Optional[str] = SAFE_SHORT_STR_VALIDATION
-    text: Optional[str] = Field(min_length=1, max_length=MAX_STR_INPUT_LENGTH)
+    modelName: Optional[str] = Field(
+        min_length=0, max_length=500, pattern=r"^[A-Za-z0-9-_. /:]*$", default=None
+    )
+    provider: Optional[str] = SAFE_SHORT_STR_VALIDATION_OPTIONAL
+    sessionId: Optional[str] = SAFE_SHORT_STR_VALIDATION_OPTIONAL
+    workspaceId: Optional[str] = SAFE_SHORT_STR_VALIDATION_OPTIONAL
+    mode: Optional[str] = SAFE_SHORT_STR_VALIDATION_OPTIONAL
+    text: Optional[str] = Field(
+        min_length=1, max_length=MAX_STR_INPUT_LENGTH, default=None
+    )
     files: Optional[List[FileFieldValidation]]
-    modelKwargs: Optional[ModelKwargsFieldValidation]
+    modelKwargs: Optional[ModelKwargsFieldValidation] = None
 
 
 class InputValidation(BaseModel):
@@ -67,11 +75,15 @@ def handler(event, context: LambdaContext):
         "userId": event["identity"]["sub"],
         "data": request.get("data", {}),
     }
-    InputValidation(**message)
 
     try:
+        InputValidation(**message)
         response = sns.publish(TopicArn=TOPIC_ARN, Message=json.dumps(message))
         return response
+    except ValidationError as e:
+        errors = e.errors(include_url=False, include_context=False, include_input=False)
+        logger.warning("Validation error", errors=errors)
+        raise ValueError(f"Invalid request. Details: {errors}")
     except Exception as e:
         # Do not return an unknown exception to the end user.
         logger.exception(e)

lib/chatbot-api/functions/resolvers/send-query-lambda-resolver/requirements.txt

@@ -1,2 +1,2 @@
-pydantic==2.4.0
+pydantic==2.9.2
 aws_xray_sdk==2.14.0

lib/rag-engines/aurora-pgvector/create-aurora-workspace.ts

@@ -137,6 +137,10 @@ export class CreateAuroraWorkspace extends Construct {
             ? cdk.RemovalPolicy.RETAIN_ON_UPDATE_OR_DELETE
             : cdk.RemovalPolicy.DESTROY,
         retention: props.config.logRetention,
+        // Log group name should start with `/aws/vendedlogs/` to not exceed Cloudwatch Logs Resource Policy
+        // size limit.
+        // https://docs.aws.amazon.com/step-functions/latest/dg/bp-cwl.html
+        logGroupName: `/aws/vendedlogs/states/CreateAuroraWorkspace-${this.node.addr}`,
       }
     );
 

lib/rag-engines/data-import/file-import-workflow.ts

@@ -111,6 +111,10 @@ export class FileImportWorkflow extends Construct {
           ? cdk.RemovalPolicy.RETAIN_ON_UPDATE_OR_DELETE
           : cdk.RemovalPolicy.DESTROY,
       retention: props.config.logRetention,
+      // Log group name should start with `/aws/vendedlogs/` to not exceed Cloudwatch Logs Resource Policy
+      // size limit.
+      // https://docs.aws.amazon.com/step-functions/latest/dg/bp-cwl.html
+      logGroupName: `/aws/vendedlogs/states/FileImportStateMachine-${this.node.addr}`,
     });
 
     const workflow = setProcessing.next(fileImportJob).next(setProcessed);

lib/rag-engines/data-import/website-crawling-workflow.ts

@@ -132,6 +132,10 @@ export class WebsiteCrawlingWorkflow extends Construct {
           ? cdk.RemovalPolicy.RETAIN_ON_UPDATE_OR_DELETE
           : cdk.RemovalPolicy.DESTROY,
       retention: props.config.logRetention,
+      // Log group name should start with `/aws/vendedlogs/` to not exceed Cloudwatch Logs Resource Policy
+      // size limit.
+      // https://docs.aws.amazon.com/step-functions/latest/dg/bp-cwl.html
+      logGroupName: `/aws/vendedlogs/states/WebsiteCrawling-${this.node.addr}`,
     });
 
     const workflow = setProcessing.next(webCrawlerJob).next(setProcessed);

lib/rag-engines/kendra-retrieval/create-kendra-workspace.ts

@@ -89,6 +89,10 @@ export class CreateKendraWorkspace extends Construct {
             ? cdk.RemovalPolicy.RETAIN_ON_UPDATE_OR_DELETE
             : cdk.RemovalPolicy.DESTROY,
         retention: props.config.logRetention,
+        // Log group name should start with `/aws/vendedlogs/` to not exceed Cloudwatch Logs Resource Policy
+        // size limit.
+        // https://docs.aws.amazon.com/step-functions/latest/dg/bp-cwl.html
+        logGroupName: `/aws/vendedlogs/states/CreateKendraWorkspace-${this.node.addr}`,
       }
     );
 

lib/rag-engines/opensearch-vector/create-opensearch-workspace.ts

@@ -151,6 +151,10 @@ export class CreateOpenSearchWorkspace extends Construct {
             ? cdk.RemovalPolicy.RETAIN_ON_UPDATE_OR_DELETE
             : cdk.RemovalPolicy.DESTROY,
         retention: props.config.logRetention,
+        // Log group name should start with `/aws/vendedlogs/` to not exceed Cloudwatch Logs Resource Policy
+        // size limit.
+        // https://docs.aws.amazon.com/step-functions/latest/dg/bp-cwl.html
+        logGroupName: `/aws/vendedlogs/states/CreateOpenSearchWorkspace-${this.node.addr}`,
       }
     );
 

lib/rag-engines/workspaces/delete-document.ts

@@ -170,6 +170,10 @@ export class DeleteDocument extends Construct {
           ? cdk.RemovalPolicy.RETAIN_ON_UPDATE_OR_DELETE
           : cdk.RemovalPolicy.DESTROY,
       retention: props.config.logRetention,
+      // Log group name should start with `/aws/vendedlogs/` to not exceed Cloudwatch Logs Resource Policy
+      // size limit.
+      // https://docs.aws.amazon.com/step-functions/latest/dg/bp-cwl.html
+      logGroupName: `/aws/vendedlogs/states/DeleteWorkspace-${this.node.addr}`,
     });
 
     const stateMachine = new sfn.StateMachine(this, "DeleteDocument", {

lib/rag-engines/workspaces/delete-workspace.ts

@@ -173,6 +173,10 @@ export class DeleteWorkspace extends Construct {
           ? cdk.RemovalPolicy.RETAIN_ON_UPDATE_OR_DELETE
           : cdk.RemovalPolicy.DESTROY,
       retention: props.config.logRetention,
+      // Log group name should start with `/aws/vendedlogs/` to not exceed Cloudwatch Logs Resource Policy
+      // size limit.
+      // https://docs.aws.amazon.com/step-functions/latest/dg/bp-cwl.html
+      logGroupName: `/aws/vendedlogs/states/DeleteWorkspace-${this.node.addr}`,
     });
 
     const stateMachine = new sfn.StateMachine(this, "DeleteWorkspace", {

lib/shared/file-import-batch-job/requirements.txt

@@ -4,12 +4,11 @@ numpy==1.26.0
 cfnresponse==1.1.2
 aws_requests_auth==0.4.3
 requests-aws4auth==1.2.3
-langchain==0.2.14
-langchain-community==0.2.12
+langchain==0.3.5
+langchain-community==0.3.3
 opensearch-py==2.3.1
 psycopg2-binary==2.9.7
 pgvector==0.2.2
-pydantic==2.4.0
 urllib3<2
 openai==1.47.0
 beautifulsoup4==4.12.2

lib/shared/index.ts

@@ -45,7 +45,7 @@ export class Shared extends Construct {
     this.kmsKeyAlias = props.config.prefix + "genaichatbot-shared-key";
     this.queueKmsKeyAlias =
       props.config.prefix + "genaichatbot-queue-shared-key";
-    const powerToolsLayerVersion = "46";
+    const powerToolsLayerVersion = "2";
 
     this.defaultEnvironmentVariables = {
       POWERTOOLS_DEV: "false",
@@ -258,10 +258,12 @@ export class Shared extends Construct {
       stringValue: JSON.stringify(props.config),
     });
 
+    //https://docs.powertools.aws.dev/lambda/python/3.2.0/
+    const pythonVersion = pythonRuntime.name.replace(".", "");
     const powerToolsArn =
       lambdaArchitecture === lambda.Architecture.X86_64
-        ? `arn:${cdk.Aws.PARTITION}:lambda:${cdk.Aws.REGION}:017000801446:layer:AWSLambdaPowertoolsPythonV2:${powerToolsLayerVersion}`
-        : `arn:${cdk.Aws.PARTITION}:lambda:${cdk.Aws.REGION}:017000801446:layer:AWSLambdaPowertoolsPythonV2-Arm64:${powerToolsLayerVersion}`;
+        ? `arn:${cdk.Aws.PARTITION}:lambda:${cdk.Aws.REGION}:017000801446:layer:AWSLambdaPowertoolsPythonV3-${pythonVersion}-x86_64:${powerToolsLayerVersion}`
+        : `arn:${cdk.Aws.PARTITION}:lambda:${cdk.Aws.REGION}:017000801446:layer:AWSLambdaPowertoolsPythonV3-${pythonVersion}-arm64:${powerToolsLayerVersion}`;
 
     const powerToolsLayer = lambda.LayerVersion.fromLayerVersionArn(
       this,

lib/shared/layers/common/requirements.txt

@@ -3,15 +3,15 @@ numpy==1.26.0
 cfnresponse==1.1.2
 aws_requests_auth==0.4.3
 requests-aws4auth==1.2.3
-langchain==0.2.14
-langchain-community==0.2.12
-langchain-aws==0.1.17
-langchain-openai==0.1.25
-openai==1.47.0
+langchain==0.3.5
+langchain-core==0.3.13
+langchain-community==0.3.3
+langchain-aws==0.2.4
+langchain-openai==0.2.4
+langchain-text-splitters==0.3.1
 opensearch-py==2.4.2
 psycopg2-binary==2.9.7
 pgvector==0.2.2
-pydantic==2.4.0
 urllib3<2
 beautifulsoup4==4.12.2
 requests==2.32.0

lib/shared/layers/python-sdk/python/genai_core/chunks.py

@@ -7,7 +7,7 @@
 import genai_core.opensearch.chunks
 from genai_core.types import CommonError, Task
 from typing import List, Optional
-from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_text_splitters import RecursiveCharacterTextSplitter
 
 PROCESSING_BUCKET_NAME = os.environ.get("PROCESSING_BUCKET_NAME", "")
 s3 = boto3.resource("s3")

lib/shared/utils.ts

@@ -23,13 +23,14 @@ export abstract class Utils {
     }
   }
 
-  static getDefaultEmbeddingsModel(config: SystemConfig): string {
+  static getDefaultEmbeddingsModel(config: SystemConfig): string | undefined {
     const defaultModel = config.rag.embeddingsModels.find(
       (model) => model.default === true
     );
 
     if (!defaultModel) {
-      throw new Error("No default embeddings model found");
+      // No default embdeding is set in the config when Aurora or Opensearch are not used.
+      return undefined;
     }
 
     return `${defaultModel.provider}::${defaultModel.dimensions}::${defaultModel.name}`;

lib/shared/web-crawler-batch-job/requirements.txt

@@ -4,11 +4,10 @@ numpy==1.26.0
 cfnresponse==1.1.2
 aws_requests_auth==0.4.3
 requests-aws4auth==1.2.3
-langchain==0.2.14
+langchain==0.3.5
 opensearch-py==2.3.1
 psycopg2-binary==2.9.7
 pgvector==0.2.2
-pydantic==2.4.0
 urllib3<2
 openai==0.28.0
 beautifulsoup4==4.12.2

lib/user-interface/react-app/src/pages/rag/embeddings/embeddings.tsx

@@ -72,10 +72,6 @@ export default function Embeddings() {
     validate: (form) => {
       const errors: Record<string, string | string[]> = {};
 
-      if (!form.embeddingsModel) {
-        errors.embeddingsModel = "Embeddings model is required";
-      }
-
       for (let i = 0; i < form.input.length; i++) {
         const input = form.input[i];
         if (input.trim().length === 0) {

lib/user-interface/react-app/vite.config.ts

@@ -10,7 +10,7 @@ export default defineConfig({
   define: {
     "process.env": {},
     // Prevents replacing global in the import strings.
-    global: "global",
+    global: isDev ? {} : "global",
   },
   plugins: [
     isDev && {